Similarly the account of Viçvāmitra’s dialogue ends with the assertion that the Bharatas successfully crossed the rivers in search of booty, having won a passage by the intercession of their priest. The interesting, but obscure, hymn (iv. 42), in which Indra and Varuṇa seem to engage in a dispute as to their relative pre-eminence, is clearly commented on by the poet himself, and his intervention may be suspected even where it is not essential. Now it is clear that the tradition of the ritual literature did not know what to make of the dialogues of the Ṛgveda. The genre of composition was one which died out in the later Vedic age; it is significant that the Atharvaveda knows but one hymn of that type (v. 11) in which the priest, Atharvan, begs the god for the payment due, a cow; the god is little inclined to accord his prayer, but finally is induced to relent and to add to the guerdon due the promise of eternal friendship. It is not in the least surprising, therefore, if we find that Yāska and Çaunaka in the fifth century B.C. were at variance as to whether the hymn x. 95 was a dialogue, as the former held, or a mere legend, as the latter believed. [4] In the commentary of Sāyaṇa we find that the tradition was unable to ascribe any ritual use for nearly all the hymns; the case of x. 86 is an exception, but it is significant that that hymn has little of a true dialogue, the three speakers rather uttering enigmas than conversing, and it was therefore easier to fit it into the inconspicuous part it occupies in the later ritual. We must, therefore, admit that we have in these dialogues the remnant of a style of poetry which died out in the later Vedic period. Its original purpose is obscure, but a very interesting suggestion was made in 1869 by Max Müller in connexion with his version of Ṛgveda i. 165. [5] He conjectured that the ‘dialogue was repeated at sacrifices in honour of the Maruts or that possibly it was acted by two parties, one representing Indra, the other the Maruts and their followers’. In 1890 the suggestion was repeated with approval by Professor Lévi, [6] who added to it the argument that the Sāmaveda shows that the art of music had been fully developed by the Vedic age. Moreover the Ṛgveda [7] already knows maidens who, decked in splendid raiment, dance and attract lovers, and the Atharvaveda [8] tells how men dance and sing to music. There is, therefore, a priori no fatal objection to assuming that the period of the Ṛgveda knew dramatic spectacles, religious in character, in which the priests assumed the rôles of gods and sages in order to imitate on earth the events of the heavens.